Platelet-derived microparticles provoke chronic lymphocytic leukemia malignancy through metabolic reprogramming

Abstract

Background: It is well established that inflammation and platelets promote multiple processes of cancer malignancy. Recently, platelets have received attention for their role in carcinogenesis through the production of microvesicles or platelet-derived microparticles (PMPs), which transfer their biological content to cancer cells. We have previously characterized a new subpopulation of these microparticles (termed mito-microparticles), which package functional mitochondria. The potential of mitochondria transfer to cancer cells is particularly impactful as many aspects of mitochondrial biology (i.e., cell growth, apoptosis inhibition, and drug resistance) coincide with cancer hallmarks and disease progression. These metabolic aspects are particularly notable in chronic lymphocytic leukemia (CLL), which is characterized by a relentless accumulation of proliferating, immunologically dysfunctional, mature B-lymphocytes that fail to undergo apoptosis. The present study aimed to investigate the role of PMPs on CLL metabolic plasticity leading to cancer cell phenotypic changes.

Methods: CLL cell lines were co-incubated with different concentrations of human PMPs, and their impact on cell proliferation, mitochondrial DNA copy number, OCR level, ATP production, and ROS content was evaluated. Essential genes involved in metabolic-reprogramming were identified using the bioinformatics tools, examined between patients with early and advanced CLL stages, and then validated in PMP-recipient CLLs. Finally, the impact of the induced metabolic reprogramming on CLLs' growth, survival, mobility, and invasiveness was tested against anti-cancer drugs Cytarabine, Venetoclax, and Plumbagin.

Results: The data demonstrated the potency of PMPs in inducing tumoral growth and invasiveness in CLLs through mitochondrial internalization and OXPHOS stimulation which was in line with metabolic shift reported in CLL patients from early to advanced stages. This metabolic rewiring also improved CLL cells' resistance to Cytarabine, Venetoclax, and Plumbagin chemo drugs.

Conclusion: Altogether, these findings depict a new platelet-mediated pathway of cancer pathogenesis. We also highlight the impact of PMPs in CLL metabolic reprogramming and disease progression.

Keywords: extracellular vesicles; leukemia; microparticles; microvesicles; mitochondria; platelets.

Figure 1

Interaction analysis of platelet-derived microparticles and CLL cells. (A) Flow cytometry was conducted for the characterization of CLL cell lines labeled with anti-human CD19 antibody either untreated (control) or co-incubated with platelet-derived microparticles (PMPs) labeled with MitoTracker Deep Red (MTDR). Co-incubations of CII (upper panels) or MEC-1 cells (lower panels) with PMPs were performed at 1:10 and 1:100 (cell/PMP) ratios for 48 hours. (B) CLL co-incubations (48 hours) with PMPs (1:10 and 1:100) were also photographed by fluorescence microscopy. CLL cell lines were labeled with DAPI and MitoTracker™ Green FM (MTG), whereas PMPs were labeled with MTDR. (C) Relative mitochondrial DNA copy number was obtained from cell/PMP co-cultures for 24 and 48 hours using Real-time PCR on leucine tRNA (mtDNA-tRNA-Leu) normalized with nuclear DNA copy number. Each value is the mean ± SEM of six separate experiments. The asterisk (*) indicates significant differences where **p < 0.01, and ***p < 0.001.

Figure 2

Impact of mitochondria acquisition on CLL viability. CLL cell lines were evaluated for cell viability at varying co-culture ratios (1:10 and 1:100) and incubation time (24 and 48 hours) with either (A) PMPs or (B) purified mitochondria isolated from the Set2 megakaryocytic cell line. Viability was established using a cell-based colorimetric assay (Cell Titer-Blue, Promega) combined a multi-well plate spectral analysis from a Synergy H1 Hybrid Multi-Mode Microplate Reader (Ex/Em: 560/590 nm). Each value is the mean ± SEM of six separate experiments. The asterisk (*) indicates significantly different (*p < 0.05, ***p < 0.001). (C) Light microscopy (40X) of CLLs co-incubations with PMPs or Set2-derived mitochondria was also performed for morphological examination.

Figure 3

Impact of PMPs on CLL metabolic functions. CLL cell lines (CII and MEC-1) were evaluated for metabolic functions at varying co-culture ratios (1:10 and 1:100) and incubation periods (24 and 48 hours) with either PMPs or purified mitochondria isolated from the Set2 megakaryocytic cell line. (A) Oxygen consumption rate (OCR) was determined using an Oroboros-O2k for: ROUTINE respiration as the basal state; the non-phosphorylating resting state (LEAK respiration); and the electron transport system (ETS) capacity. (B) Total ATP and glycolytic ATP levels were evaluated using a firefly luciferase-based assay. Relative ATP concentrations related to specific pathways (oxidative phosphorylation/OXPHOS versus glycolysis) were determined by cell treatments to inhibit mitochondrial complexes I and III using Rotenone (ROT) and Antimycin A (AmA), respectively. (C, D) Cells were treated with AmA, ROT, and phorbol 12-myristate 13-acetate (PMA) as ROS inducers. Thereafter, samples were mixed with (C) CellROX green reagent and SYTOX Red Dead Cell Stain prior analysis by flow cytometry at Ex/Em: 508/525 nm (CellROX green) and 640/658 nm (SYTOX Red Dead Cell Stain) to determine intracellular total ROS content; or with (D) MitoSOX™ reagent prior analysis by flow cytometry at Ex/Em: 510/580 nm to determine mitochondrial superoxide content. N-acetylcysteine (NAC) and Tert-butyl hydroperoxide (TBHP) were used as negative and positive controls, respectively. Results are expressed as the mean ± SEM of six biological experiments. One-way ANOVA followed by Tukey’s multiple comparisons test show significant differences in the values presenting different superscript letters (p < 0.05).

Figure 4

Bioinformatics identification and analysis of core metabolic-related genes involved in CLL progression. (A) Identification of key genes involved in CLL metabolism. The interaction between genes was visualized by the Cytoscape plugins BisoGenet and CentiScaPe, and subsequently annotated using the EnrichR. (B) Expression analysis of the most differentially expressed metabolic-related genes (DEGs) between CLL datasets were then calculated and plotted in relative expression. (C) Pearson Chi-square test of CLL samples was then analyzed to plot the expression levels of TFAM and NRF1 as mitochondrial biogenesis indicators in relation to Binet stages of clinical samples. (D) The most active signaling pathways from patients with early onset of CLL (Binet stage A), or (E) advanced disease (CLL Binet stage C) were also identified and listed. Each value is the mean ± SEM of six separate experiments. The asterisk (*) and number sign (#) indicate significantly different from the control group (non-PMP treated), and the 1:10 PMP-treated group, respectively (p <.05).

Figure 5

Impact of PMPs on CLL chemoresistance and metabolic gene expression. (A) CII and (B) MEC-1 cell lines were co-incubated with PMPs (1:100, 48 h) and then treated with increasing amounts of either Cytarabine, Venetoclax, and Plumbagin (5-200 nM) for 24 hours to establish the lethal doses for 50% mortality (LD50). Control groups were treated with DMSO as dilution buffer. Light microscopy of cells treated at their respective LD50 are also presented for morphological analysis. (C, D) Real-time PCR analysis was performed to evaluate differentially expressed metabolic-related genes (DEGs) in CLL groups in the presence of chemo-drugs where each column represents the PMP-treated/non-treated cell ratio. Control groups were treated with DMSO as dilution buffer. Results are expressed as the mean ± SEM of six biological experiments. One-way ANOVA followed by Tukey’s multiple comparisons test show significant differences in the values presenting different superscript letters (p < 0.05).

Figure 6

Effects of PMPs on CLL survival and apoptosis. CLL cells were co-incubated with PMPs (1:100, 48 h) prior to treatments with the cells’ respective LD50 of Cytarabine, Venetoclax, and Plumbagin for 24 hours for the evaluation of drug cytotoxic events. Flow cytometry analysis in (A) CII and (B) MEC-1 cells stained with Propidium Iodide (PI, X‐axis) and Annexin‐FITC (X‐axis) sorted cell populations according to distinctive quarter zones (Q), including the large nuclear fragments/debris or Q1 (Anx^-/PI⁺), necrosis cells or Q2 (Anx⁺/PI⁺), apoptotic cells or Q3 (Anx⁺/PI^-), and the living cells or Q4 (Anx^-/PI^-). Values for each quadrant were then plotted in relation to their respective control groups (DMSO). Each value is the mean ± SEM of six separate experiments. The asterisk (*) indicates significantly different (*p < 0.05, ***p < 0.001).

Figure 7

Impact of PMPs on CLL cycle progression. CLL cells were co-incubated with PMPs (1:100, 48 h) prior to treatments with the cells’ respective LD50 of Cytarabine, Venetoclax, and Plumbagin for 24 hours then evaluated for cycle analysis using flow cytometry. Cell accumulation in each cell cycle phase (i.e., G2/M) for (A) CII and (B) MEC-1 cells stained with Propidium Iodide and RNase A was then plotted in relation to control groups treated with DMSO. Each value is the mean ± SEM of six separate experiments. The asterisk (*) indicates significantly different (*p < 0.05, **p < 0.01, ***p < 0.001).

Figure 8

Effect of PMPs on CLL malignant processes. CLL cells were co-incubated with PMPs (1:100, 48 h) prior to treatments with the cells’ respective LD50 of Cytarabine, Venetoclax, and Plumbagin for 24 hours then evaluated for (A) migration and (B) invasion capacity using 24-well ThinCerts. Control groups were treated with DMSO as dilution buffer. Light microscopy of invading cells treated at their respective LD50 are also presented for morphological analysis (left panels). Results are expressed as the mean ± SEM of 6 biological experiments. One-way ANOVA followed by Tukey’s multiple comparisons test show significant differences in the values presenting different superscript letters (p < 0.05).